Targeted gene modification in the mouse (commonly referred to as knockout mouse technology because the goal of many of the modifications is to abolish, or knock out, target gene function) is the most effective method for discovery of mammalian gene function in live animals and for creating genetic models of human disease. Knockout mouse creation typically begins by introducing a targeting vector into mouse embryonic stem (ES) cells. The targeting vector is a linear piece of DNA comprising a selection or marker gene (e.g., for drug selection) flanked by mouse DNA sequences—the so-called homology arms—that are similar or identical to the sequences at the target gene and which promote integration into the genomic DNA at the target gene locus by homologous recombination. To create a mouse with an engineered genetic modification, targeted ES cells are introduced into mouse embryos, for example pre-morula stage (e.g., 8-cell stage) or blastocyst stage embryos, and then the embryos are implanted in the uterus of a surrogate mother (e.g., a pseudopregnant mouse) that will give birth to pups that are partially or fully derived from the genetically modified ES cells. After growing to sexual maturity and breeding with wild type mice some of the pups will transmit the modified gene to their progeny, which will be heterozygous for the mutation. Interbreeding of heterozygous mice will produce progeny that are homozygous for the modified allele and are commonly referred to as knockout mice.
The initial step of creating gene-targeted ES cells is a rare event. Only a small portion of ES cells exposed to the targeting vector will incorporate the vector into their genomes, and only a small fraction of such cells will undergo accurate homologous recombination at the target locus to create the intended modified allele. To enrich for ES cells that have incorporated the targeting vector into their genomes, the targeting vector typically includes a gene or sequence that encodes a protein that imparts resistance to a drug that would otherwise kill an ES cell. The drug resistance gene is referred to as a selectable marker because in the presence of the drug, ES cells that have incorporated and express the resistance gene will survive, that is, be selected, and form clonal colonies, whereas those that do not express the resistance gene will perish. Such a selectable marker is typically present in a selection cassette, which typically includes nucleic acid sequences that will allow for expression of the selectable marker. Molecular assays on drug-resistant ES cell colonies identify those rare clones in which homologous recombination between the targeting vector and the target gene results in the intended modified sequence (e.g., the intended modified allele).
After selection of drug-resistant clones, the selection cassette typically serves no further function for the modified allele. Ideally the cassette should be removed, leaving an allele with only the intended genetic modification, because the selection cassette might interfere with the expression a neighboring gene such as a reporter gene, which is often incorporated adjacent to the selectable marker in many knockout alleles, or might interfere with a nearby endogenous gene (see, e.g., Olsen et al, (1996) Know Your Neighbors: Three Phenotypes of the Myogenic bHLH Gene MRF4. Cell 85:1-4; Strathdee et al. (2006) Expression of Transgenes Targeted to the Gt(ROSA)26Sor Locus Is Orientation Dependent, PloS ONE 1(1):e4). Either event can confound the interpretation of the phenotype of the modified allele. For these reasons selectable markers in knockout alleles are usually flanked by recognition sites for site-specific recombinase enzymes, for example, loxP sites, which are recognized by the Cre recombinase (see, e.g., Dymecki (1999) Site-specific recombination in cells and mice, in Gene Targeting: A Practical Approach, 2d Ed., 37-99). A typical selection cassette comprises a promoter that is active in ES cells linked to the coding sequence of an enzyme, such as neomycin phosphotransferase, that imparts resistance to a drug, such as G418, followed by a polyadenylation signal, which promotes transcription termination and 3′ end formation and polyadenylation of the transcribed mRNA. This entire unit is flanked by recombinase recognition sites oriented to promote deletion of the selection cassette upon the action of the cognate recombinase.
Recombinase-catalyzed removal of the selection cassette from the knockout allele is typically achieved either in the gene-targeted ES cells by transient expression of an introduced plasmid carrying the recombinase gene or by breeding mice derived from the targeted ES cells with mice that carry a transgenic insertion of the recombinase gene. Either method has its drawbacks. Selection cassette excision by transient transfection of ES cells is not 100% efficient. Incomplete excision necessitates isolating multiple subclones that must be screened for loss of the selectable marker, a process that can take one to two months and subject a targeted clone to high levels of recombinase and a second round of electroporation and plating that can adversely affect the targeted clone's ability to transmit the modified allele through the germline. Consequently, the process might require repetition on multiple targeted clones to ensure the successful creation of knockout mice from the cassette-deleted clones.
The alternative approach of removing the selection cassette in mice requires even more effort. To achieve complete removal of the selection cassette from all tissues and organs, mice that carry the knockout allele must be bred to an effective general recombinase deletor strain. But even the best deletor strains are less than 100% efficient at promoting cassette excision of all knockout alleles in all tissues. Therefore, progeny mice must be screened for correct recombinants in which the cassette has been excised. Because mice that appear to have undergone successful cassette excision may still be mosaic (i.e., cassette deletion was not complete in all cell and tissue types), a second round of breeding is required to pass the cassette-excised allele through the germline and ensure the establishment of a mouse line completely devoid of the selectable marker. In addition to about six months for two generations of breeding and the associated housing costs, this process may introduce undesired mixed strain backgrounds through breeding, which can make interpretation of the knockout phenotype difficult.
Accordingly, there remains a need in the art for compositions and methods for excising nucleic acid sequences in genetically modified cells and animals.